Σημασιολογική ανάκτηση νομικών κειμένων
Semantic retrieval of legal documents
View/
Keywords
Τεχνητή νοημοσύνη ; Επεξεργασία φυσικής γλώσσας ; Δικαστικές υποθέσεις ; Ανάκτηση πληροφορίας ; Ομοιότητα κειμένων ; Νομικά κείμεναAbstract
In a variety of legal settings, there is a clear need to use a text as a query, in order to retrieve related documents
. For example, in customary law, each case has related cases in the past that the lawyers and judges need to consult. In other applications, a complaint or lawsuit is related to specific laws or decisions. Natural Language Processing can support semantic text similarity, while Information Retrieval can help in retrieving the related documentss. This project will touch the intersection of the two domains, trying to build efficient representations, comparison methods and indexes to facilitate semantic-relevance-based document retrieval in the legal domain. Using the above known techniques in the legal field, we examine which is more appropriate in this field.
In this work, more emphasis is placed on the evaluation part. To evaluate the techniques we have implemented, we proposed a method of human evaluation. As part of this method, we created a annotation tool, so that the process we proposed could be implemented. The purpose was to do experiments with annotators, to see how appropriate the techniques we used for this field were and which method is the most suitable.
Postgraduate Studies Programme
Τεχνητή Νοημοσύνη - Artificial IntelligenceDepartment
Σχολή Τεχνολογιών Πληροφορικής και Επικοινωνιών. Τμήμα Ψηφιακών ΣυστημάτωνCorporate Department
Εθνικό Κέντρο Έρευνας Φυσικών Επιστημών «Δημόκριτος»Number of pages
64Language
GreekCollections
Except where otherwise noted, this item's license is described as
Αναφορά Δημιουργού 3.0 Ελλάδα
Αναφορά Δημιουργού 3.0 Ελλάδα